Quantitative Analysis of Ki67, a Prognostic Biomarker in Gastroenteropancreatic Neuroendocrine Tumors

Abstract

Summary:

1. Ki67 Labeling Index (LI) is the preferred method of establishing grade for GEP-NETs
2. Several approaches to measuring Ki67 LI exist in practice
3. Machine learning offers an attractive method for calculating Ki67 LI
4. Pathologists need to be accurate and precise when performing pixel-based labeling of training data
5. While strict guidelines for Ki67 LI in GEP-NETs do not exist, expert opinion and analogy to other tissues is helpful
6. KiNet is a single-stage deep-learning-based detection and classification pipeline with performance at or above the state of the art